Model Upload Using OFSAA Data Model Descriptor (Database.XML) File

This feature allows you to resume data model upload from the logical data model in the form of OFSAA Data Model Descriptor file (Database.XML) that is generated in a base environment. This helps in speeding up the model upload process, by skipping the XSL transformation in the primary environment. This feature can be used in case the same model in the development environment needs to be uploaded to multiple OFSAA instances in the production environment. In such scenarios, you can copy the model definition (database.XML) files and scripts to the target environment and run the command line utility CopyUpload.sh, to integrate those files in the target environment. You can choose to resume model upload process from script generation or script execution.

Following are the steps involved:

1.      Copy the required files from source to the target environment, based on from where you want to resume the model upload process.

2.     Execute the CopyUpload utility.

3.     Perform the model upload.

Copying the Required Files

Based on the selection of your start point, copy the required files from your source environment to the desired location.

1.      If the start point is script generation, copy the <INFODOM>_DATABASE.XML file from /ftpshare/<INFODOM>/erwin/fipxml/ folder on the source.

2.     If the start point is script execution, copy the <INFODOM>_DATABASE.XML from /ftpshare/<INFODOM>/erwin/fipxml/ folder as well as the DB scripts from /ftpshare/<INFODOM>/erwin/scripts and /ftpshare/<INFODOM>/scripts folders.

 

Start point

Required Files

Script generation

/ftpshare/<INFODOM>/erwin/fipxml/<INFODOM>_DATABASE.xml 

Script Execution

/ftpshare/<INFODOM>/erwin/fipxml/<INFODOM>_DATABASE.xml

DB Scripts from /ftpshare/<INFODOM>/erwin/scripts and /ftpshare/<INFODOM>/scripts folders

 

Executing CopyUpload Utility

The command line utility CopyUpload is used to prepare the target OFSAA instance to resume the model upload process from script generation or script execution. This utility is available in the $FIC_HOME/ficapp/common/FICServer/bin/ folder.

Following are the prerequisites for executing the utility:

·       CopyUpload.sh should have Execute permissions.

·       Appropriate permissions should be granted on the source folders.

·       All the required files should have been copied to the target environment. For details, see Copying the Required Files.

To run the utility from the console:

1.      Navigate to $FIC_HOME/ficapp/common/FICServer/bin.

2.     Execute the utility using the command:

./CopyUpload.sh

3.     Enter the following when prompted:

§       Enter ftpshare location- the path of the ftpshare location

§       Enter dsnname – the information domain name

§       Enter absolute filepath of database xml - the {, llc;x’ path of the folder in which the <INFODOM>_DATABASE.XML file is available

§       Continue with scripts transfer? [y,n]– Enter ‘y’ if you want to copy scripts also, else enter ‘n’.

§       Enter absolute path for table folder– the path of the folder in which the table is available.

§       Enter absolute path for alter table– the path of the folder in which the alter table file is available

§       Enter absolute path for scripts– the path of the folder in which the DB scripts are available

4.    Once the utility is executed successfully, the files are copied to the following locations:

§       //ftpshare/archive/<INFODOM>/Erwin/fipxml/<INFODOM>_DATABASE.xml

§       //ftpshare/archive/<INFODOM>/Erwin/scripts/

§       //ftpshare/archive/<INFODOM>/scripts

Triggering Model Upload

Trigger the model upload process either through command line or through UI.

 

NOTE

CopyUpload.sh should have been executed successfully

 

To perform model upload using Data Model Descriptor:

1.      From the Business Model Upload window, select Upload Option as Data Model Descriptor.

2.     Select the Object Registration Mode from the drop-down list as Full Object Registration or Incremental Object Registration. It is recommended to select incremental only if the changes are minimal.

 

NOTE

Incremental Object Registration should be opted only if the object registration on the base environment was incremental. Full Object Registration can be performed irrespective of mode opted in the base environment.

 

3.     Select the Use archived database xml check box.

4.    Select the Use archived scripts check box if the starting point of model upload process is from script execution. That is, if you have copied the DB scripts to the target environment. Otherwise, deselect the check box.

5.     Select either Yes / No to directly Update the database schema with Model changes.

§       If you select Yes, the generated SQL scripts are executed at runtime to update the Model changes in the database.

§       If you select No, it restricts the system from updating the database automatically with Model changes and only the model scripts will be created. Later you should execute the SQL scripts in a correct sequence, in order to make the Infodom Schema to be consistent with the DATABASE.xml. For more information, see Sequence of Execution of Scripts section.

Also when you select No, ensure the following:

§       You have a third party tool or ETL tool to manage the schema updates.

§       Database consistency and schema updates are maintained manually by the database administrator.

 

NOTE

The table scripts are only created and needs to be updated manually. If you choose this option for the first time and later perform an Incremental / Sliced / Complete Model Re-build, you need to manually synchronize the schema with the database schema.

 

6.    Select Yes for the Generate DDL execution logs option if you want execution audit information such as execution start time, end time, and status of each SQL statements run as part of the model upload process. The execution log file will be available under ftpshare/<INFODOM>/Erwin/executionlogs folder.

7.     Select Yes for Refresh Session Parameters option to use Database session parameters during model upload process. For more information, see Configuring Session Parameters section.

8.    Select either Yes / No to directly update the Alter constraints in NOVALIDATE state. During incremental or sliced model upload, alterations to constraints consumes a lot of time as the constraints need to be validated.

§       If you select Yes, an option to alter the constraints in NOVALIDATE state is enabled and it will not check the existing data for the integrity constraint violation. It is quite useful in cases where it is known that the existing data is clean. So, NOVALIDATE can potentially reduce the additional overhead of the constraint validation and it would enhance the performance.

§       By default the option is selected as No. If you select No, then the option to alter the constraints is not enabled and it will check the existing data for the integrity constraint violation.

 

NOTE

Note the following points about NOVALIDATE option.

·       Constraints in NOVALIDATE state are supported only in incremental and sliced modes.

·       Model upload process irrespective of the status of success or failure will bring the constraints into NOVALIDATE state. Hence, ENABLE VALIDATE should be done as a post-model upload activity. That is, Rollback does not validate the constraints which are non-validated during the upload activity.

·       NOVALIDATE option is not relevant for HDFS systems.

 

9.    Click Upload Model.